Abstract:

In order to overcome the deficiency of Renyi entropy-based thresholding method on images with inherent fuzzy characteristics， a valid multilevel thresholding method based on fuzzy set theory is presented. The Renyi entropies of fuzzy partitions are defined when the image is converted to the fuzzy domain， and then the image is segmented by maximum entropy principle. In addition， the differential evolution algorithm is used to search the optimal thresholds and improve the computational efficiency in implementation of the proposed method. Experimental results on both synthetic and real images demonstrate the effectiveness of the proposed method. Meanwhile， the proposed method satisfies the time performance requirements of image segmentation when the differential evolution algorithm is used.